There is a pressing need for research on the implementation and evaluation of interventions designed to reduce medication errors in "real world" ambulatory care settings. Medication errors in ambulatory care are common, yet the vast majority of research has been conducted in inpatient settings. Based on our previous research and clinical experience, this application will address these issues by refining and experimentally evaluating a novel pharmacy alert system that intercepts medication errors after a prescription has been written but before a pharmacist dispenses the drug. We will conduct three separate randomized controlled trials to evaluate the impact of the intervention on the occurrence of medication errors for: 1) 4,000 pregnant women, 2) 19,000 patients with chronic kidney disease, and 3) 21,000 patients who receive high-risk drugs requiring laboratory monitoring. Randomization at the patient level will assign participants to either a "usual care" group or an "intervention" group. The primary outcome of the study will be the occurrence of medication errors. The intervention's potential for translation and generalizability using the RE-AIM framework of reach, effectiveness, adoption, implementation, and maintenance will also be evaluated. The proposed project will advance scientific knowledge in the field of patient safety. Most medication error reduction strategies have focused on the point of prescribing and have been evaluated in the inpatient setting. The proposed project is focused in the point of dispensing and will be implemented and evaluated in the outpatient setting. The proposed project will be population based (it will be conducted at all 19 pharmacies and will include all eligible members of a large health plan) and will assess the true effectiveness (as opposed to efficacy) of the intervention. Pharmacists working in standard clinical settings and under routine circumstances will deliver the intervention. As a result, the study should also provide a realistic estimate of intervention effects in other settings.